SGOTM 14 - One Short Straw

[shyuhe]

An exploring WB would be good for trying to find foreign trade routes as well. Not sure what the soonest is for us to spare hammers for that though.

 

[ZPV]

Big Picture Strategy

I suppose no one has tested out this scenario to the end. I haven't. It might not be a bad idea for one of our faster players to do that.

I'm wondering (as always) if the fastest conquest route is an Astro beeline. With no circumnavigation, this time galleons double our movement over galleys in addition to saving extra turns by going directly across ocean tiles. Furthermore, we're philosophical, so we can generate the GSes even faster. Of course, the galleons are also much safer against galleys and triremes and carry 50% more units.

I've noticed that by the time we get our first GS, the academy or super spec only saves us about a turn on math. Our capital is not particularly commerce rich. I'm wondering if we even want to use our GSes that way at all. What if instead we grew our empire to 1AD and had Astro at the same time?

The risk with an astro beeline is, what if we have to face a highly-promoted tough unit (the wizard) to win? We can't do an astro beeline + conquest and tech up to a unit strong enough to kill it - the only way numbers might work is to use flanking II elephants thanks to that promotion's immunity to first strikes.

The other thing to note is, land war *is* possible on hub maps. They certainly manage ok on quick speed on multiplayer. We just need to be prepared to commit several workers to building (rail?)roads down the isthmi, and we'll be able to get our units there plenty quick enough.

Galleons can work - the outer ring of tiles on the map are usually ocean - but I'd be surprised if they're the answer for more than the odd coastal city we want to catch pants-down.

 

[LowtherCastle]

Here's the challenge, oh ye mighty naysayers. ZPV, play the new test save out to 1000BC, assuming we want GLH and Mids. Then I'll test it out my way, someone else test it out whatever way you think is best. This is the challenge: COme within 50t of my finish time...

 

[shyuhe]

... 50 turns??? Did you find some uber GE generating world builder trick?

 

[Dhoomstriker]

Let's say that the Wizard is a Drill VI, City Garrison III, Combat VI, etc Machine Gun or Rifleman

Is it really impossible to kill it with masses of weaker units? If we threw 70 Axemen at it, wouldn't it die? Or, as was suggested, Flanking II units (which are easy to get with a Barracks + Vassalage or a Barracks + a Stable) out of every City, to avoid the Dril Promos... if we threw 70 Flanking II Chariots at it, wouldn't it die? Either that or like 40 Combat II Horse Archers, since they are inherrently immune to First Strikes?

Okay, I threw up a simple test:
-> 0% Cultural Defences, since I think that it is fair that even if the Wizard is in a City, we can send one measely Catapult to Bomard its defences down to 0%
-> 25% Fortification bonus
-> City on a Hills square
-> Two units: A Rifleman in one City and a Machine Gun in another City
-> Each unit had: up to and including Drill VI, City Garrison III, Combat VI, Guerrilla III, Formation (anti-Horse-units) plus some other random promotions like up to Medic III. Note that I, of course, gave Drill I, Drill II, Drill III, etc for all prior promotions, although it would be easy enough to accidentally forget to do so when using the World Builder, which is why I mention that point
-> It was easiest to just attack using non-promoted Horse Archers, so those are what I used

It took 22 Horse Archers to kill the Rifleman and 18 Horse Archers to kill the Machine Gun. Obviously, we'd need to plan to attack with more units in case of bad random generation luck, but those numbers should give you a rough idea that advanced units may not be necessary.

Imagine, even, that it is a Machine Gun (I think a Rifleman is "fairer" given what I am about to say). Teams that play an average game will almost certainly have Riflemen before the game is over. However, Machine Guns get an inherrent bonus against Gunpowder units, so using Riflemen might not be the best choice of attackers anyway. I would not be surprised if it were a Rifleman unit, since then Grenadiers could kill it. We have to assume that neilmeister made the map winnable by average-level teams, and therefore I would think that a unit like a Rifleman makes sense, since the average player doesn't consider the idea of bringing 50+ units in an attack but would not think it unreasonable to bring 10 Grenadiers in an attack.

However, just in case it is a Mech Infantry, I'll go and run my test again... hang on...
An unpromoted (but fully Fortified) Hills-City-defending Mech Infantry without extra promos died to 35 Horse Archers. However, when I gave it all of the promos that could possibly help it, it took 116 Horse Archers to take it down.

With an unpromoted, Hills-defending Modern Armour unit (it can't get a fortification bonus), it took 34 Horse Archers. However, when I gave it all of teh promos that could possibly help it, it took 140 Horse Archers.


I think that those numbers sound feasible, considering that:
a) if we have to face a tougher unit, so will other teams
b) we will almost certainly have Cuirassiers instead by that point in the game, which would be overkill relative to the Horse Archers
c) Is it really likely that neilmeister gave SUCH a tough unit? Even if he did, then we'd need far less than 200 Horse Archers to take out the toughest unit possible in the game

 

[ZPV]

I think that those numbers sound feasible, considering that:
a) if we have to face a tougher unit, so will other teams
b) we will almost certainly have Cuirassiers instead by that point in the game, which would be overkill relative to the Horse Archers
c) Is it really likely that neilmeister gave SUCH a tough unit? Even if he did, then we'd need far less than 200 Horse Archers to take out the toughest unit possible in the game

We don't have horses. We can't bet on having horses. Elephants are the only unit we have available that can get flanking 2, and without immunity to first strikes we could suicide a hundred weak units against a highly promoted unit without making a scratch.
It's a big, expensive, stack of elephants with otherwise useless promotions that can take down the uber-rifle.

 

[ZPV]

Here's the challenge, oh ye mighty naysayers. ZPV, play the new test save out to 1000BC, assuming we want GLH and Mids. Then I'll test it out my way, someone else test it out whatever way you think is best. This is the challenge: COme within 50t of my finish time...

You're on! Just let me put in a suitably-promoted wizard

 

[LowtherCastle]

You're on! Just let me put in a suitably-promoted wizard

Right. And make it according to your predictions, based on the health bars. Plus, add to the city all possible buildings that aren't wonders that could provide additional defense (walls, castles, etc.), including any civ-specific UBs.

 

[LowtherCastle]

Okay, considering Dhoom's test data, what are the possible ways we could attempt to conquer 5 of the 8 AIs in preparation for a domination victory?

1. WEs + cats + roads/galleys. After GLH+Mids, we beeline Construction + HBR*** and have at it.
2. Discover iron => swordsmen + cats + r/gs. Requires constr+IW.
3. Discover iron or copper => macemen + cats +r/gs. Requires CS+Mach+constr.
4. Discover horses => HAs + cats + r/gs. Requires HBR + constr.
5. Discover horses + iron => knights + cats +r/gs. Requires HBR + constr + guilds.
6. musketmen + cats + r/gs. Requires GP + constr.
7. Grenadiers + cats + r/gs. Requires Mil Sci.
8. LBMs + cats + r/gs. Requires Feudalism + constr.

Any other options less advanced than grens?

***Dhoom reminded me...

 

[shyuhe]

I believe you can field an army of catapults and LB without strategic resources. It's pretty sub-optimal though.

 

[Mitchum]

I just tried to match ZPV's GLH date in Marble and failed by 3 turns. Did you whip the LH to transfer hammers from the LH to the GLH, ZPV? I didn't see any whip unhappiness anywhere but it most likely wore off...

 

[ZPV]

I just tried to match ZPV's GLH date in Marble and failed by 3 turns. Did you whip the LH to transfer hammers from the LH to the GLH, ZPV? I didn't see any whip unhappiness anywhere but it most likely wore off...

I 2-pop whip both the LH and GLH with the 3-turn delay in settling marble.

 

[Dhoomstriker]

We don't have horses. We can't bet on having horses. Elephants are the only unit we have available that can get flanking 2, and without immunity to first strikes we could suicide a hundred weak units against a highly promoted unit without making a scratch.
It's a big, expensive, stack of elephants with otherwise useless promotions that can take down the uber-rifle.

Okay, let's run the math with War Elephants.

War Elephants are far more cost-effective than Horse Archers when you don't need the Horse Archers' extra movement... the only possible downside is the time that it takes to march 1-movement units to the right part of the map... so, we'd probably build stacks of War Elephants first then the last few units could be Horse Archers (assuming that there is at least one Horse Resource on the map), just to "catch up" to the War Elephants... otherwise, we'll just go with pure War Elephants plus whatever other units we have nearby at the time.


Costs
Horse Archer = 50 Hammers
War Elephant = 60 Hammers
Maceman = 70 Hammers


The Setup
This time, I even upped the ante by surrounding the Barb Cities with Rivers.

I'm attaching my test map.

Unless I made a mistake, every Barb unit is:
a) On a Hills square
b) Surrounded by Rivers
c) Promoted up to and including Drill VI, City Garrison III, Guerilla III, Combat VI, Formation (anti-horse units)
d) Surrounded by stacks of 20 War Elephants each equipped with Flanking II


The Results
I used "random seed on reload" and ran the tests 3 times per Barb unit type. Listed below are the number of War Elephants required to kill the unit for each of the 3 test runs.
Barb Rifileman = 16, 14, 22
Barb Machine Gun = 26, 25, 27
Barb Mech Infantry = 50, 58, 60
Barb Modern Armour = 62, 69, 59


Feasibility
Flanking II should be easy enough to get with Barracks and Vassalge or Stables.

If we build or capture The Pyramids, we could optionally use Police State, but we could certainly use Slavery to create a massive army in a short period of time if it is the end of the game and we don't care about keeping our Cities in good condition.

The numbers are even more favourable than for Horse Archers, although I admit that I used unpromoted Horse Archers instead of Combat II Horse Archers. That said, the Horse Archers didn't have to go up against River defenses, so the comparison to the Horse Archer results won't be exact, but given the relative costs, as long as we can walk the War Elephants that we build within a reasonable timeframe, then War Elephants should be more than sufficient to take down a single Barb unit, no matter how uber he is.


A note on Stack Attacks

Spoiler :

Note that if you care to repeat the results, I suggest enabling Stack Attack and then selecting units as a group. Stack Attack only attacks with the units that you have selected, so, if you only attack with 2 units out of 20 in a stack selected, then only those 2 units will attack, but they will both attack "at the same time" such that if the first one wins the battle, the second unit will "move" into the City alongside of the first unit. I.e. Stack Attack commits all of the selected units to moving and if they are moving into a square where enemy units exist, then those units are also all committed to attacking (in an often silly manner when you have multiple unit types selected, which is why it's generally a good option to keep disabled).

 

[Mitchum]

@ZPV Based on the length of the health bars, what are you able estimate the Wizard to be? My guess is that it is NOT a Mech Infantry or Modern Armour, but Neil could have been evil...

 

[LowtherCastle]

@ZPV Based on the length of the health bars, what are you able estimate the Wizard to be? My guess is that it is NOT a Mech Infantry or Modern Armour, but Neil could have been evil...

It's consistent with something about as powerful as a rifle existing at turn 0. It would also behave like this, if neil had placed a rifle on the map, and then deleted it.
This might be an AI, though - there was at least one with power higher than expected. Possibly a witch; or else a way to increase the cost of the first or second war.
edit: promotions don't affect the health bars, so it could be CG3 C6 G3 W3 March Drill4...

 

[LowtherCastle]

I assume there's no way for us to currently determine the barbs power (or their starting power)?

 

[LowtherCastle]

The risk with an astro beeline is, what if we have to face a highly-promoted tough unit (the wizard) to win? We can't do an astro beeline + conquest and tech up to a unit strong enough to kill it - the only way numbers might work is to use flanking II elephants thanks to that promotion's immunity to first strikes.

The other thing to note is, land war *is* possible on hub maps. They certainly manage ok on quick speed on multiplayer. We just need to be prepared to commit several workers to building (rail?)roads down the isthmi, and we'll be able to get our units there plenty quick enough.

Galleons can work - the outer ring of tiles on the map are usually ocean - but I'd be surprised if they're the answer for more than the odd coastal city we want to catch pants-down.

I believe the ~14 AI cities built so far have been predominantly coastal, no?

One of the huge advantages of naval travel is that our ships aren't slowed by their coastal water culture. Another is chaining. Roads don't really compete well at all. It surely seems likely that this map is closer to archipelago than pangaea on the continuum. A third advantage is our movement from one hub to the next. Galleons will be 1-2 turns away rather than 5 or 10 turns for roads or galleys.

 

[ZPV]

Here's the challenge, oh ye mighty naysayers. ZPV, play the new test save out to 1000BC, assuming we want GLH and Mids. Then I'll test it out my way, someone else test it out whatever way you think is best. This is the challenge: COme within 50t of my finish time...

Here you are. The barb unit is a redcoat with a jillion first strikes and first strike chances from promotions.

OK. Does this mean we should actually try to develop Stone (i.e. net the crab and grow to size 3 or 4) as we build the 'Mids? Or do you still want to stagnate at 1 pop working the stone quarry but use fewer workers to do the chopping?

ZPV, can you give me a quick rundown of how you built the GLH in 1080 BC? I'm thinking something like this:

- Steal corn from Washington and grow
- Build library until Sailing comes in and switch to LH -> GLH
- Work netted crab at 2 pops to keep growing.
- Work mined marble at 3 pops
- Work mined GHill at 4 pops
- Chop 3 forests into GLH once Math is learned

It needs two whips. The worker mining will finish the mine and then improve marble.
Steal the corn immediately.
Grow to size 4 working sensible tiles; whip LH.
Revolt to slavery 1880. The library in Washington gets built after two turns of 0% (the worker from gems will go to the plains hill forest in Wash. to chop).
Grow to size 4 again, optimize for hammers, get three chops by the end of t72. Whip t73.

 

[LowtherCastle]

Using Dhooms save, I tried unpromoted grenadiers against the rifleman. First 20 died without a single HP of damage. Reload. 50 then died, knocking the rifle down to 11.2/14HP. Meanwhile, 12 measly WEs killed the rifle with 8 casualties and 3 retreats...

Those drill promotions pack a punch.

 

[Mitchum]

While you guys work on killing the Wizard, I'll put a PPP together through Sailing (7 turns). Most of the "stuff" suggested by ZPV above will happen after this turnset.